1. Field of the Invention
The invention relates to a solid-state imaging device, a method of manufacturing the same, and an electronic apparatus such as a camera including the solid-state imaging device.
2. Description of the Related Art
A digital video camera or digital still camera which is for use by consumers has called for high resolution power to transmit details of a photographic subject and a reduced-size device with a regard to portability. In addition, in order to realize these demands, development on a reduction in the pixel size while maintaining the image capturing property has been performed in regard to a solid-state imaging device (image sensor). However, in recent years, in addition to existing demands for high resolution and reduction in size, demand has been increasing for improvement in low luminance for photographic subjects or high-speed image capturing and the like. In order to realize these, expectation has increased for improvement in comprehensive image quality starting from the SN ratio in the solid-state imaging device.
The CMOS solid-state imaging device is categorized into a front-illuminated device shown in FIG. 5 and a back-illuminated device shown in FIG. 6. The front-illuminated solid-state imaging device 111 includes a pixel region 113 where a plurality of unit pixels 116 composed of photodiode PDs, which become a photoelectric conversion section, and a plurality of pixel transistors, are formed in plural number on a semiconductor substrate 112, as shown in the schematic configuration diagram of FIG. 5. The pixel transistor (not shown) represents a gate electrode 114 in FIG. 5, representing schematically the presence of the pixel transistor. Each of the photodiodes PD is isolated by an element isolation region 115 formed of an impurity diffused layer. A multilayer of interconnection layers 119, in which a plurality of interconnections 118 is disposed through an interlayer insulating film 117, is formed on the surface side of a semiconductor substrate 112 where the pixel transistor is formed. The interconnection 118 is formed in other parts than those corresponding to the position of the photodiode PD. On the multilayer of the interconnection layers 119, an on-chip color filter 121 and an on-chip microlens 122 are formed in this order through a planarization film 120. The on-chip color filter 121 is constituted by arranging each color filter of, for example, red (R), green (G) and blue (B). In the front-illuminated solid-state imaging device 111, the light L is incident from the side of the substrate surface, on which the multilayer of the interconnection layers 119 is formed, using this substrate surface as a light receiving surface 123.
A back-illuminated type solid-state imaging device 131 includes the pixel region 113 where a plurality of unit pixels 116 composed of the photodiodes PD, which become a photoelectric conversion section, and a plurality of pixels transistors are formed in plural number on the semiconductor substrate 112, as shown in the schematic configuration diagram of FIG. 6. The pixel transistor (not shown) is formed on the surface side of the substrate, and represents a gate electrode 114 in FIG. 6, representing schematically the presence of the pixel transistor. Each of the photodiodes PD is isolated by the element isolation region 115 formed of an impurity diffused layer. The multilayer of the interconnection layers 119, in which a plurality of interconnections 118 is disposed through the interlayer insulating film 117, is formed on the surface side of the semiconductor substrate 112 where the pixel transistor is formed. In the back-illuminated device, the interconnection 118 can be formed regardless of the position of the photodiode PD. On the other hand, an insulating layer 128, the on-chip color filter 121 and the on-chip microlens 122 are formed in this order on the backside of the semiconductor substrate 112 to which the photodiode PD faces. In the back-illuminated solid-state imaging device 131, the light L is incident from the substrate backside, which is the side opposite to the substrate surface on which the multilayer of the interconnection layers 119 and the pixel transistor are formed, using this substrate backside as a light receiving surface 132. Since the light L is incident to the photodiode PD with no restriction of the multilayer of the interconnection layers 119, apertures of the photodiode PD can be broadly taken, and high sensitivity can be achieved.
The present applicants have succeeded in the development of experimental production of a back-illuminated CMOS solid-state imaging device that improves sensitivity, which is one of the main elements for high image quality and noise reduction rate by changing the basic structure of pixels to the back-illuminated type without losing the advantages of low power consumption and high speed that the CMOS solid-state imaging device has. This developed back-illuminated CMOS solid-state imaging device has 5 million effective pixels, each pixel size being 1.75 μm×1.75 μm, and is driven at a rate of 60 frames per second.
In the front-illuminated device of a related art, the interconnection 118 or the pixel transistor on the surface side of a substrate, where the photodiodes PD are formed, hinders the incident light collected with an on-chip microlens, which is an issue in reduction of pixel size and variation of incident angle. In comparison to this, in the back-illuminated device, applying the light from the backside inverted from the silicon substrate allows an increase in the amount of the light incident to unit pixels while also suppress sensitivity reduction in regard to angle variation of incident light with no influence of the interconnection 118 or the pixel transistor.
The back-illuminated CMOS solid-state imaging device is disclosed, for example, in Japanese Unexamined Patent Application Publication Nos. 2003-31785, 2005-353631, 2005-353955, and 2005-347707. In addition, a technique of using hafnium oxide (HfO2) as an antireflection film used in the back-illuminated CMOS solid-state imaging device is disclosed in Japanese Unexamined Patent Application Publication No. 2007-258684.
Solid-state imaging devices are largely divided into a CCD (Charge Coupled Device) type solid-state imaging device and a CMOS (Complementary Metal Oxide Semiconductor) type solid-state imaging device.
In these solid-state imaging devices, the light-receiving portion composed of photodiodes is formed for each pixel. In the light-receiving portion, signal charges are generated by photoelectric conversion by incident light to the light-receiving portion. In the CCD type solid-state imaging device, signal charges generated in the light-receiving portion are transferred to a charge transfer portion that has the CCD structure and output in the output portion as converted to pixel signals. On the other hand, in the CMOS type solid-state imaging device, signal charges generated in the light-receiving portion are amplified for each pixel and the amplified signals are output as pixel signals by signal ray.
In such solid-state imaging device, there are problems that aliasing is generated in a semiconductor substrate by tilted incident light or incident light diffusely reflected in the upper portion of the light-receiving portion, and optical noise such as smear, flare is generated.
Japanese Unexamined Patent Application Publication No. 2004-140152 mentioned below describes the constitution of the CCD type solid-state imaging device that allows the suppression of smear generation by forming a light-shielding film, which is formed in the upper portion of the charge transfer portion, to be buried in the groove portion that is formed in the interface between the light-receiving portion and the read gate portion. Since the CCD type solid-state imaging device in Japanese Unexamined Patent Application Publication No. 2004-140152 is constituted to form the light-shielding film in the groove portion formed using the LOCOS oxide film, it is difficult to form the light-shielding film deep in the substrate, and not possible to completely prevent incidence of tilted light, which is a cause of smear. In addition, since the pixel area is reduced in proportion to the burying depth of the light-shielding film, it is practically difficult to bury deep the light-shielding film.
In recent years, a back-illuminated solid-state imaging device has been proposed, in which the light is applied from the side opposite to the side of the substrate on which the interconnection layer is formed (see Japanese Unexamined Patent Application Publication No. 2004-71931 described below). In the back-illuminated solid-state imaging device, the light application side does not include the interconnection layer, the circuit element and the like, and thereby the aperture rate of the light-receiving portion formed on the substrate can be enhanced. Furthermore, since the incident light is incident to the light-receiving portion with no reflecting to the interconnection layer and the like, sensitivity is improved.
Even in such a back-illuminated solid-state imaging device, there is a concern about optical noise due to tilted light, and thus a light-shielding film is preferably formed between the light-receiving portions of the backside of the substrate, which becomes the light application side. In this case, a layer, that has the light-shielding film on the backside of the substrate that becomes the light application side, may be considered to be formed as one layer. However, the distance between the substrate and the on-chip lens side is lengthened in proportion to the height of the light-shielding film, and thus deterioration of light-collecting property may occur. When the light-collecting property is deteriorated, problems may occur such that the tilted light transmitted through a color filter of other pixels is incident to the light-receiving portion of different pixels from the pixels, and color mixing and sensitivity reduction are also generated.
Meanwhile, it has been found that in the back-illuminated CMOS solid-state imaging device, the light-collecting structure made of only the on-chip microlens 122 has following problems remarkably that may happen.
(1) It is very difficult to completely suppress optical color mixing with adjacent pixel. It may not be a problem in use such as monitoring, cellular phone, but color mixing has to be further reduced in use of audio/video (AV) (camcorder, digital still camera and the like).
(2) A light-shielding film is provided in the effective pixel peripheral portion to prevent noise in the peripheral circuit region and determine optical black level. However, the light-collecting state changes in the effective pixel peripheral portion by level difference of the light-shielding film, and thus uniform optical property is not realized. That is, as shown in FIG. 4, the light-shielding film 126 is formed through the insulating film 127 from optical black level region (so-called optical black region) 113B outside the effective pixel region 13A to the peripheral circuit section 125. On top of this, the on-chip color filter 121 and the on-chip microlens 122 are formed. At this time, height difference d of the lens surface of the on-chip microlens 122 occurs by the level difference by presence or absence of the light-shielding film 126, in the peripheral portion of the effective pixel region 113A and the central part in the inside thereof. The light-collecting state changes due to this height difference d, and the peripheral portion becomes dark in comparison to the brightness of the central part of the effective pixel region, and thus uniform optical property is not obtained. So-called sensitivity unevenness happens.
(3) Reflection happens by the on-chip microlens 122 or the on-chip color filter 121 in photographing using high-intensity light source. Diffracted light is reflected to the seal glass and the like on the package of the solid-state imaging device, and further incident thereto, and color mixing happens uniformly to the RGB pixel. By this color mixing, streaky image defect of Mg color (hereinafter, called as flare of Mg color) happens in a radial fashion from the high-intensity light source, which is unique in the back-illuminated solid-state imaging device.
Specifically, the problems will be described using the green pixel 151G and the red pixel 151R of FIG. 3A. The light L that is incident to the on-chip microlens 122 of the green pixel 151G, is incident to the photodiode PD of the green pixel through the green filter 121G. However, some tilt light La is incident to the photodiode PD of the red pixel 151R that is adjacent to the pixel boundary. This is shown with simulation of the light intensity when the light of 550 nm wavelength is incident to two pixels of the green pixel and the red pixel of FIG. 3B. In FIG. 3B, the regional part A represents a part where the light intensity is strong, the light-colored regional part B represents a part where the light intensity is weak, and the heavy-colored regional part (streaky part) C represents a part where the light intensity is almost absent. The fine periodical streak pattern shows the progress of the light wave surface. If the light is looked at, which is incident to the photodiode PD under the light receiving surface 153, it is found that weak light is incident to the photodiode PD of the red pixel 151R, and color mixing happens in the region D shown in the round shape near the pixel boundary.
On the other hand, as shown in FIG. 1, a seal glass 135 is disposed through a space 134 in the window of the incident light side of the package (not shown) where the back-illuminated CMOS solid-state imaging device 131 is stored. Furthermore, an optical low-pass filter 136 is disposed on this seal glass 135 through the space 134, and an infrared cutoff filter 137 is disposed on this through the space 134. Furthermore, a camera lens 138 is disposed on the upper side. The incident light L1, which transmits the camera lens 138 and incident to the solid-state imaging device 131, is reflected partially at each medium interface of the solid-state image capturing element 131. The incident light L1 is mainly reflected at the lens surface of the on-chip microlens 122, and the silicon surface that becomes the light receiving surface. The on-chip microlens 122 is periodically arranged, and thus diffraction phenomenon occurs. The reflected and diffracted light L2, which has been reflected in the solid-state imaging device 131, is reflected in various angles such as nearly vertical reflection, distant-direction reflection. The light L2 is reflected at the seal glass 135, the optical low-pass filter 136 and the infrared cutoff filter 137, and is incident again to the solid-state imaging device 131 as re-incident light L3. Among them, the light diffracted at great angle is reflected at the seal glass 135, and incident again to the solid-state imaging device 131, which becomes Mg flare 141 in a radial fashion shown in FIG. 2 (see the circular frame E). The white streak (white flare) 142 in a radial fashion occurs due to the iris in camera lens side, and is a phenomenon that also occurs in the front-illuminated solid-state imaging device, and cause no such uncomfortable feeling. However, Mg flare 141, which is unique in the back-illuminated solid-state imaging device, is noticeable in comparison to the back green, for example, when photographing is taken for sunbeams streaming through leaves, and becomes problems.
Happening of this Mg flare 141 is due to the treatment of the white balance in the process of the signal processing for uniform spectroscopic characteristics of red (R), green (G) and blue (B). Although color mixing is performed for each pixel to be equal by re-incidence of the diffracted light, Mg flare happens since signals of red (R) and blue (B) come to have greater gain in comparison to that of green (G), and thus are emphasized with the white balance treatment.
In the back-illuminated solid-state imaging device, Mg flare may happen by above-described optical color mixing and reflected light by leakage of incident light to adjacent pixel. However, optical color mixing to adjacent pixel may happen also in the front-illuminated solid-state imaging device.